JP2010193213A - Hearing aid - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve a problem wherein an S/N ratio is not improved and the hearing recognition degree of a hearing aid user is not improved unless not only the volume of an input sound signal is amplified but also only a desired signal is extracted by discriminating the desired signal from a disturbing signal in order to improve the hearing recognition degree of the hearing aid user; and, when a complicated discrimination process using multiple parameters is used for the analysis of a surrounding environment, a new issue is developed in terms of previous learning or long-time drive. <P>SOLUTION: A noise level is calculated from an input sound signal, and a stationary noise suppression/non-stationary noise suppression are changed over from each other in accordance with the calculated noise level, and thereby the type of noise to be suppressed and the number of sound sources are controlled instead of strength of noise suppression. Thereby, a certain S/N ratio for a hearing-impaired person is always kept from a quiet scene to a noisy scene only by a small number of parameters such as a noise level, and this hearing aid allowing a user to easily hear sound can be provided at any time. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a hearing aid that suppresses noise according to a noise level.

  A hearing aid is a device that is used by a hearing-impaired person, a person whose hearing has declined, or the like to assist hearing. A hearing aid converts an externally generated acoustic signal into an electrical signal using a microphone, amplifies the level of this electrical signal, converts it back to an acoustic signal with a receiver such as an earphone, and so on as an audible sound that can be heard by the user. Output.

  As for the acoustic signal output from the receiver, in addition to sound information necessary for daily life such as conversation voice, output sound of TV, radio, etc., intercom and telephone bell, etc., users such as living noise and environmental noise are required. There are various disturbing sounds that interfere with the recognition of sound information.

  Therefore, conventionally, in hearing aid processing of a hearing aid, a low-level sound is amplified and a high-level sound is not amplified. There have been various devices that combine amplification and suppression to make it easier for users to hear.

  Particularly in recent years, digital hearing aids that convert acoustic signals input from a microphone into digital signals and perform the above-mentioned hearing aid processing by digital processing have become mainstream. In such a digital hearing aid, the input acoustic signal is divided into a plurality of bands, and the desired signal / interfering signal (for example, voice / non-voice) for each band is processed at high speed to extract only the desired signal. This system is equipped with processing such as directional sound collection that extracts only acoustic signals coming from the front using advanced noise suppression and input time differences from microphones installed in two places before and after the hearing aid. Furthermore, hearing aids of a type that have a storage area inside the hearing aid, hold a plurality of hearing aid processing programs, and automatically switch the hearing aid processing to the hearing aid user's usage scene and surrounding environment, or the user manually switches the hearing aid are becoming popular. is there.

  Many concepts have been proposed for switching hearing aid programs according to the user's surrounding environment. For example, Patent Document 1 proposes a technique for holding a plurality of input / output characteristic conversion tables for determining an output signal level corresponding to an input signal level and switching the conversion table as appropriate according to the frequency characteristic of the input signal ( FIG. 6). Patent Document 2 proposes a method using auditory scene analysis (ASA: Auditory Scene Analysis) or HMM (Hidden Markov Model: Hidden Markov Model) for analysis of the surrounding environment (FIG. 7).

  A conventional hearing aid 1001 shown in FIG. 6 converts an acoustic signal input from a microphone 1002 into a digital signal by an A / D converter 1003, performs hearing aid processing by an input / output characteristic conversion unit 1004, and then performs a D / A converter 1005. This is a hearing aid of the type that is converted into an acoustic signal again and output from the receiver 1006. The input / output characteristic conversion unit 1004 is configured by a table storing the converted amplitude value to be output using the amplitude value of the input digital signal as an address, and is configured to control the output signal level according to the input signal level. Yes. The conversion table applied to the input / output characteristic conversion unit 1004 is defined by the input / output characteristic determination unit 1007, and the input / output characteristic determination unit 1007 should switch the conversion table to be applied according to the characteristics of the input acoustic signal. Holds multiple conversion tables. For the determination of the conversion table, the user designates from the outside with a switch or the like, and the degree of environmental noise can be evaluated using an input signal and can be automatically selected and switched.

A conventional hearing aid 2001 shown in FIG. 7 is a type of hearing aid in which an acoustic signal input from microphones 2002 and 2003 is subjected to hearing aid processing by a hearing aid processing unit 2004 and output from a receiver 2005. The hearing aid 2001 performs feature extraction on the input acoustic signal by the signal analysis unit 2006, identifies the instantaneous acoustic environment situation of the acoustic signal by the signal identification unit 2007, and performs hearing aid processing according to the identified acoustic environment situation. The hearing aid algorithm of the part 2004 is switched. The instantaneous acoustic environment state in the signal identification unit 2007 is identified by a combination of auditory base features such as sound intensity, spectral form, and harmonic structure, and HMM or the like is applied as an identification algorithm. HMM is a statistical method widely used in speech recognition and the like, and is a model for estimating an output state for an unknown input from past state transitions and appearance probability distributions in each state. In addition, a training device 2008 is provided to appropriately initialize parameters so as not to fall into local optimization.
Japanese Patent No. 2568134 JP-T-2004-500592 Publication

  However, in the input / output characteristic conversion table of the above-mentioned patent document 1, the desired signal buried in the disturbing signal is extracted and S is extracted even though the amplification of the disturbing signal section is limited and the ease of hearing and the uncomfortable feeling are reduced. / N is not improved. Therefore, it does not lead to an improvement in intelligibility or hearing recognition. For a hearing aid user, it can be said that the S / N improvement in the hearing aid processing is an indispensable requirement for improving the hearing recognition.

  In addition, the input / output characteristic conversion table of the above-mentioned Patent Document 1 is held finely for each frequency band that is divided into bands, or the technology using the HMM as in the above-mentioned Patent Document 2 is very complicated and burdensome. Will be expensive.

  Hearing aids often use small air batteries as a drive power source for the purpose of use, and it is desired that they can be used continuously for about one to two weeks with a single replacement. Battery replacement of a small device such as a hearing aid is very troublesome for an elderly person who is the main hearing aid user, and reducing the frequency of battery replacement by driving for a long time is also an important functional element for the hearing aid. Therefore, in the hearing aids of Patent Document 1 and Patent Document 2, since the load is high, the batteries must be frequently replaced.

  The HMM method disclosed in Patent Document 2 is an identification method that is widely used in speech recognition and the like, and the identification process is very heavy. In addition, it is indispensable to construct a database by pre-learning in order to make the HMM function effectively. Pre-learning to adapt to various user environments and securing a storage area that holds a huge beta base Is hard to say.

  The present invention solves the above-mentioned conventional problems, and is an adaptive hearing aid that automatically switches the hearing aid processing according to the surrounding environment, and is a simple and high-speed adaptive type adapted to the hearing characteristics and purpose of use of the hearing impaired The object is to provide a hearing aid.

  In order to achieve the above object, a hearing aid according to the present invention includes a noise detector that calculates a noise level of an ambient environment from an input acoustic signal captured by a microphone, and a stationary noise that removes stationary noise from the input acoustic signal. And a non-stationary noise suppression unit that removes non-stationary noise from the input acoustic signal, the steady noise suppression unit functions when the noise level is equal to or lower than a predetermined threshold, and the noise level is The non-stationary noise suppression unit functions when the predetermined threshold value is exceeded. In addition, a directivity control unit that generates a signal having directivity sensitivity in a specific direction from input acoustic signals captured by at least two microphones, and at least one of the input acoustic signals is used to control the noise level of the surrounding environment. A noise detector for calculating, a stationary noise suppressing unit for removing stationary noise from the input acoustic signal, and an unsteady noise suppressing unit for removing unsteady noise from the input acoustic signal, The directivity control unit generates a non-directional signal that does not have directivity sensitivity in a specific direction, and the steady noise suppression unit functions with the non-directional signal as an input, and the noise When the level is equal to or higher than a predetermined threshold, the directivity control unit generates a directivity signal having directivity sensitivity in a specific direction, and the unsteady noise suppression unit functions with the directivity signal as an input. I am characterized in.

  The present invention can be realized not only as such a hearing aid, but also as an integrated circuit, a hearing aid processing method, a program for causing a computer to execute the hearing aid processing, and a recording medium for storing the program. .

  According to the hearing aid of the present invention, it is possible to remove the disturbing sound that reduces the hearing recognition of the hearing impaired person according to the ambient noise level, and to maintain a constant S / N from a quiet scene to a noisy scene, It can always be safe and easy to hear.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(Embodiment 1)
FIG. 1 is a configuration diagram showing a schematic configuration of a hearing aid 1 according to Embodiment 1 of the present invention. The hearing aid 1 includes a microphone 2, an amplification unit 3, a receiver 4, a noise level detection unit 5, a comparator 6, and a noise suppression unit 7. The noise suppression unit 7 further includes a steady noise suppression unit 71 and an unsteady noise suppression unit 72. The general hearing aid 1 removes the interference signal from the acoustic signal input from the microphone 2 by the noise suppression unit 7 so that only the desired signal is obtained, and the amplification unit 3 performs amplification processing according to the auditory characteristics of the user. The desired signal is amplified to the audible level and output from the receiver 4.

  Furthermore, in the hearing aid 1 of the present embodiment, the noise level is calculated by the noise level detection unit 5 using the acoustic signal input from the microphone 2. The calculated noise level signal is compared with a predetermined threshold value Th by the comparator 6, and when the noise level is less than the threshold value Th, the steady noise suppression unit 71 is caused to function as the function of the noise suppression unit 7, and the input sound Remove stationary noise from the signal. On the other hand, when the noise level is equal to or higher than the threshold Th, the hearing aid 1 causes the non-stationary noise suppression unit 72 to function as the function of the noise suppression unit 7 and removes the non-stationary noise from the input acoustic signal. In addition, by using an A / D or D / A converter not shown in FIG. 1 at the subsequent stage of the microphone 2 or the preceding stage of the receiver 4 or other appropriate location, and using digital signal processing for internal hearing aid processing, More accurate interference signal removal can be realized.

  The noise level can be calculated by substituting the input sound signal level itself, but if necessary, a combination of low-pass filter, high-pass filter, or band-pass filter and processing for cutting out the required band can be combined. good. For example, if the noise level is desired to be close to the user's hearing, a filter simulating human auditory characteristics may be applied. In addition, when it is desired to suppress the temporal fluctuation of the noise level, it may be averaged for a short time using a low-pass filter.

  FIG. 2 is a configuration diagram illustrating an example of the configuration of the noise level detection unit 5. The noise level detection unit 5 includes an auditory characteristic filter 51, an absolute value calculation unit 52, a low-pass filter 53, and a logarithmic conversion unit 54.

  In the noise level detection unit 5, the auditory characteristic filter 51 extracts an audio signal band from the input acoustic signal from the microphone 2. The absolute value calculation unit 52 takes the absolute value of the extracted audio signal band. The low-pass filter 53 smoothes a value obtained by taking the absolute value of the audio signal band. Finally, the logarithmic conversion unit 54 converts the noise level calculated by the logarithmic process of the smoothed value into a logarithmic scale [dBA].

  A typical example of steady noise suppression is a spectral subtraction method.

  FIG. 3 is a diagram showing a detailed configuration of the steady noise suppression unit 71 using the spectral subtraction method. The stationary noise suppression unit 71 includes a frequency conversion unit 711, a subtraction unit 712, a frequency inverse conversion unit 713, a noise estimation unit 714, and a voice detection unit 715. The steady noise suppression unit 71 converts the input acoustic signal into a frequency spectrum signal by the frequency conversion unit 711 and simultaneously extracts a speech section in the input acoustic signal by the speech detection unit 715. The noise estimation unit 714 estimates the noise spectrum included in the frequency spectrum signal using the extracted speech section information. The noise spectrum is estimated by, for example, smoothing a frequency spectrum signal in a section (non-voice section) other than the extracted voice section. The subtractor 712 subtracts the estimated noise spectrum from the frequency spectrum signal, and finally the frequency inverse converter 713 converts the frequency spectrum signal into an acoustic signal again.

  Similarly, a two-input spectral subtraction method effective for suppressing unsteady noise has been proposed.

  FIG. 4 is a diagram showing a detailed configuration of the unsteady noise suppression unit 72 using the two-input spectrum subtraction method. The unsteady noise suppression unit 72 includes frequency conversion units 711a and 711b, a subtraction unit 712, a frequency inverse conversion unit 713, a noise estimation unit 716, and a transfer characteristic calculation unit 717. The microphone 2 includes a first microphone 2a and a second microphone 2b. The unsteady noise suppression unit 72 converts an input sound signal mainly containing a large amount of desired signals into a frequency spectrum signal by a frequency conversion unit 711a and an input sound signal mainly containing a large amount of interference signals into a frequency spectrum signal respectively. . Each of the converted frequency spectrum signals is input to the transfer characteristic calculation unit 717, and the transfer characteristic between the two signals is calculated based on the difference information. The noise estimation unit 716 estimates a noise spectrum included in a frequency spectrum signal obtained by frequency-converting the desired signal using the calculated transfer characteristic. Since the spectrum subtraction method estimates the noise spectrum from the smoothing process of the spectrum signal, it can suppress only stationary noise with little time fluctuation, whereas the two-input spectrum subtraction method obtains the instantaneous noise spectrum from the transfer characteristics between the two inputs. By estimating, it is possible to suppress unsteady noise that fluctuates over time. The subtracting unit 712 subtracts the estimated noise spectrum from the frequency spectrum signal obtained by frequency-converting the desired signal, and finally the frequency inverse converting unit 713 converts the frequency spectrum signal into an acoustic signal again.

  In daily life, many of the noises that hinder hearing of hearing aid users, such as the noise of driving a car and the noise of crowds, are non-stationary noise that fluctuates over time. By removing not only stationary noise but also non-stationary noise, the effect of removing interfering signals contained in the input acoustic signal is significantly improved. On the other hand, for example, in an indoor living space where the surrounding environment is quiet, there are not many disturbing sounds that hinder the user's hearing in the first place, and suppressing the unsteady noise generated makes it impossible to grasp the surrounding situation. Rather, it is dangerous in terms of safety. On the other hand, stationary noise such as background noise and fan noise is often annoying even in a quiet place. When the noise level is small, it is effective to allow only steady noise suppression to function as a noise suppression method.

  When the noise level is high, for example, in crowds in the city, the volume of individual noise sources is also increasing, but it is heard that there are multiple noise sources as well as the volume that reduces the hearing perception of the hearing impaired. Is making it difficult. Speaking of unsteady indoor noise, there were many cases where it was sound information necessary for daily living such as intercoms and telephone bells, but outdoors, there were multiple environmental noises such as traffic noise, construction noise, and BGM. Makes it difficult to hear the sound you want to hear. In most cases, the sound that the hearing aid user wants to hear is the voice of the conversation partner. Therefore, when the noise level rises, it is possible to improve the degree of hearing recognition by operating unsteady noise suppression to maintain S / N and removing environmental noise other than voice.

  In the present embodiment, the spectral subtraction method / two-input spectral subtraction method are given as representative examples of steady noise suppression / unsteady noise suppression, respectively. However, in implementing the present invention, noise suppression methods are limited. It is not a thing. In addition, switching between steady noise suppression / unsteady noise suppression does not have a separate suppression processing unit, but a configuration that switches to a function that suppresses only steady noise by switching some parameters of the unsteady noise suppression method. Also good.

(Embodiment 2)
FIG. 5 is a configuration diagram showing a schematic configuration of the hearing aid 1 according to Embodiment 2 of the present invention. The hearing aid 1 includes a first microphone 2a, a second microphone 2b, an amplification unit 3, a receiver 4, a noise level detection unit 5, a comparator 6, a noise suppression unit 7, and a directivity control unit 8. Is provided. The noise suppression unit 7 further includes a steady noise suppression unit 71 and an unsteady noise suppression unit 72.

  The hearing aid 1 generates a signal having directivity sensitivity in a specific direction, which is the directivity control unit 8, using acoustic signals input from at least the first and second microphones 2 a and 2 b. At the same time, the noise level is calculated by the noise level detection unit 5 using an acoustic signal input from one or both of the first and second microphones 2a and 2b. Next, the hearing aid 1 compares the calculated noise level signal with a predetermined threshold Th in the comparator 6. When the noise level is less than the threshold Th, the directivity control unit 8 generates a non-directional signal having no directivity sensitivity in a specific direction, and at the same time, the noise suppression unit 7 performs steady noise suppression (steady noise suppression unit). 71) is made to function. On the other hand, when the noise level is equal to or higher than the threshold Th, the directivity control unit 8 generates a directivity signal having directivity sensitivity in a specific direction, and the noise suppression unit 7 performs unsteady noise suppression (unsteady noise suppression). The unit 72) is made to function to remove the interference signal included in the input acoustic signal. The hearing aid 1 amplifies the acoustic signal from which the interference signal is removed and becomes only the desired signal by the amplification unit 3 to an audible level of the user by an amplification process according to the auditory characteristic of the user, and outputs it from the receiver 4. Similarly to the first embodiment, an A / D or D / A converter is arranged at the rear stage of the first and second microphones 2a and 2b, the front stage of the receiver 4, or other appropriate place, and the internal hearing aid processing is performed. By using digital signal processing, it is possible to realize more accurate interference signal removal. The difference from the first embodiment is that at least two input microphones are required and a directivity control unit is added. Since the switching between steady noise suppression / unsteady noise suppression and the other configurations and operations are the same as those in the first embodiment, a duplicate description is omitted.

  Similar to the suppression of unsteady noise, by providing the input acoustic signal with directional sensitivity in a specific direction, the effect of removing interference signals included in other directions is significantly improved. On the other hand, in consideration of user safety, as in the suppression of non-stationary noise, in a quiet surrounding, for example, in an indoor living space, it is possible to grasp the surrounding situation by giving directional sensitivity only in a specific direction. It becomes an obstacle.

  When the noise level is high, for example, in a busy street in the city, having directivity sensitivity in a specific direction can improve the hearing recognition level by removing environmental noise existing in other directions. Furthermore, the combination of directivity control and unsteady noise suppression makes it possible to remove interference signals with higher accuracy.

  It is also possible to divide the threshold value Th in the comparator 6 into two values and perform switching between non-directional / directivity and steady noise suppression / unsteady noise suppression step by step. This has the effect of reducing the characteristic difference between hearing aid sounds due to switching, and by enabling the directivity and unsteady noise suppression step by step, especially when the interference signal removal effect of directivity + unsteady noise suppression is high. Thus, a high interference signal removal effect can be obtained without a sense of incongruity.

(Other variations)
Although the present invention has been described based on the above embodiment, it is needless to say that the present invention is not limited to the above embodiment. The following cases are also included in the present invention.

  (1) A part or all of the constituent elements constituting each of the above devices may be constituted by one system LSI (Large Scale Integration). The system LSI is an ultra-multifunctional LSI manufactured by integrating a plurality of components on a single chip, and specifically, a computer system including a microprocessor, ROM, RAM, and the like. . A computer program is stored in the RAM. The system LSI achieves its functions by the microprocessor operating according to the computer program.

  (2) A part or all of the constituent elements constituting each of the above devices may be constituted by an IC card that can be attached to and detached from each device or a single module. The IC card or the module is a computer system including a microprocessor, a ROM, a RAM, and the like. The IC card or the module may include the super multifunctional LSI described above. The IC card or the module achieves its function by the microprocessor operating according to the computer program. This IC card or this module may have tamper resistance.

  (3) The present invention may be the method described above. Further, the present invention may be a computer program that realizes these methods by a computer, or may be a digital signal composed of the computer program.

  In the present invention, the computer program or the digital signal may be recorded on a computer-readable recording medium. Examples of the computer-readable recording medium include a flexible disk, a hard disk, a CD-ROM, an MO, a DVD, a DVD-ROM, a DVD-RAM, a BD (Blu-ray Disc), and a semiconductor memory. Further, the present invention may be the digital signal recorded on these recording media.

  In the present invention, the computer program or the digital signal may be transmitted via an electric communication line, a wireless or wired communication line, a network represented by the Internet, a data broadcast, or the like.

  The present invention may be a computer system including a microprocessor and a memory, wherein the memory stores the computer program, and the microprocessor operates according to the computer program.

  In addition, the program or the digital signal is recorded on the recording medium and transferred, or the program or the digital signal is transferred via the network or the like, and executed by another independent computer system. It is good.

  (4) The above embodiment and the above modifications may be combined.

  The hearing aid of the present invention is useful as an adaptive hearing aid that automatically switches the hearing aid processing according to the surrounding environment.

It is a block diagram of the hearing aid in Embodiment 1 of this invention. It is a block diagram which shows an example of a structure of a noise detection part same as the above. It is a block diagram which shows an example of a structure of the steady noise suppression part using the spectrum subtraction method same as the above. It is a block diagram which shows an example of a structure of the unsteady noise suppression part using the 2 input spectrum subtraction method same as the above. It is a block diagram of the hearing aid in Embodiment 2 of this invention. It is a block diagram of the conventional hearing aid. It is a block diagram of the other conventional hearing aid.

DESCRIPTION OF SYMBOLS 1 Hearing aid 2 Microphone 2a 1st microphone 2b 2nd microphone 3 Amplification part 4 Receiver 5 Noise level detection part 6 Comparator 7 Noise suppression part 71 Steady noise suppression part 72 Unsteady noise suppression part 8 Directivity control part 51 Auditory characteristic Filter 52 Absolute value calculation unit 53 Low-pass filter 54 Logarithmic conversion unit 711 Frequency conversion unit 711a First frequency conversion unit 711b Second frequency conversion unit 712 Subtraction unit 713 Frequency inverse conversion unit 714, 716 Noise estimation unit 715 Audio detection unit 717 Transfer characteristic calculator

Claims (2)

A noise detector that calculates the noise level of the surrounding environment from the input acoustic signal captured by the microphone;
A stationary noise suppression unit for removing stationary noise from the input acoustic signal;
An unsteady noise suppression unit for removing unsteady noise from the input acoustic signal;
With
The steady noise suppression unit functions when the noise level is less than a predetermined threshold, and the unsteady noise suppression unit functions when the noise level is equal to or higher than the predetermined threshold;
Hearing aid characterized by. A directivity control unit that generates a signal having directivity sensitivity in a specific direction from input acoustic signals captured by at least two microphones;
Using at least one of the input acoustic signals, a noise detector for calculating a noise level of the surrounding environment;
A stationary noise suppression unit for removing stationary noise from the input acoustic signal;
An unsteady noise suppression unit for removing unsteady noise from the input acoustic signal;
With
When the noise level is less than a predetermined threshold, the directivity control unit generates an omnidirectional signal having no directivity sensitivity in a specific direction, and the steady noise suppression unit functions with the omnidirectional signal as an input. When the noise level is equal to or higher than a predetermined threshold, the directivity control unit generates a directivity signal having directivity sensitivity in a specific direction, and the unsteady noise suppression is performed using the directivity signal as an input. That the department functions,
Hearing aid characterized by.

Images